Amendments to a previous article, this piece offers updated information about the Savonius Wind Generator System.
Here are a few comments and corrections pertaining to my article The Savonius Super Rotor in the March/April 1974 issue of MOTHER EARTH NEWS.
 In case you were wondering, that's not me with the shirt off in that Image Gallery picture . . . that's David House (resident and founder of Earthmind and author of the book Methane Systems ), as photographed late in November. Yep, it were cold, but then we are real hardy folk, too.
 My comparison of the Stuart mill and the Savonius rotor omitted any mention of feathering: twisting of a conventional windplant's propeller blades to lower the device's rotational speed in high winds. (See Marcellus Jacobs: Wind-Power Generating Inventor for a discussion of the subject — MOTHER.)
Feathering is a necessary precaution in wind-charger design for a number of reasons. First — whenever a fixed gear ratio is used, some such governing arrangement is needed to prevent the alternator/generator from exceeding its maximum rated output (current). In a propeller-driven unit, however, the prop itself is a factor in determining at what windspeed feathering must occur. The problems involved may be described in terms of balance, structural design and blade tip stresses.
[a] Balance — as I mentioned in my article — is always more important (or critical) at higher rotational speeds than at lower rpm's. If a propeller is allowed to exceed its "operating range," therefore, dangerous vibrations can be set up.
[b] The structural design of a propeller, and the materials of which it's made, also determine the device's upper rotational speed beyond which centrifugal force will pull it apart.
[c] Tip stresses occur at higher windspeeds . . . but are rather independent of whether or not the propeller is turning. This is one reason why "braking" a fan to a stop in a tempest won't help it very much.
All these problems of propeller-driven units are irrelevant to the Savonius rotor. Since the Savonius device never achieves high rotational speed, balancing is not critical and centrifugal force won't pull the installation apart. In addition, a runaway Savonius rotor will soon begin to spoil itself because its own wings get in one another's way at higher rpm's.
 I'd like to recommend an additional source of information on battery maintenance: A good booklet entitled The Storage Battery (Lead-Acid): Fundamental Principles, Operation, and Care.
The table in the previous article contains a number of errors. First, the figures are for an 11 mph average annual windspeed . . . not 8 mph, as I indicated. Second, the column heading "Normal output (in watts)" should have read " Maximum normal output in watts", since the data represent highest wind situations.
The following revised chart is a more accurate indication of output at various windspeeds. Note that this information is correct only if the wind generator — or array of generators — is capable of delivering the higher wattages.
Windspeed (in mph) Output (in watts)
Here at Earthmind we're working on improved "wings", and our information package includes instructions for an alternative model which allows interchange of the wind-catching mechanism without lowering the entire assembly . . . for the benefit of folks who wish to upgrade their machines when the new information becomes available.